11 #include "pthread_impl.h"
14 #define malloc __libc_malloc
15 #define calloc __libc_calloc
16 #define realloc __libc_realloc
17 #define free __libc_free
19 /* The following is a threads-based implementation of AIO with minimal
20 * dependence on implementation details. Most synchronization is
21 * performed with pthread primitives, but atomics and futex operations
22 * are used for notification in a couple places where the pthread
23 * primitives would be inefficient or impractical.
25 * For each fd with outstanding aio operations, an aio_queue structure
26 * is maintained. These are reference-counted and destroyed by the last
27 * aio worker thread to exit. Accessing any member of the aio_queue
28 * structure requires a lock on the aio_queue. Adding and removing aio
29 * queues themselves requires a write lock on the global map object,
30 * a 4-level table mapping file descriptor numbers to aio queues. A
31 * read lock on the map is used to obtain locks on existing queues by
32 * excluding destruction of the queue by a different thread while it is
35 * Each aio queue has a list of active threads/operations. Presently there
36 * is a one to one relationship between threads and operations. The only
37 * members of the aio_thread structure which are accessed by other threads
38 * are the linked list pointers, op (which is immutable), running (which
39 * is updated atomically), and err (which is synchronized via running),
40 * so no locking is necessary. Most of the other other members are used
41 * for sharing data between the main flow of execution and cancellation
44 * Taking any aio locks requires having all signals blocked. This is
45 * necessary because aio_cancel is needed by close, and close is required
46 * to be async-signal safe. All aio worker threads run with all signals
47 * blocked permanently.
53 struct aio_thread *next, *prev;
61 int fd, seekable, append, ref, init;
64 struct aio_thread *head;
74 static pthread_rwlock_t maplock = PTHREAD_RWLOCK_INITIALIZER;
75 static struct aio_queue *****map;
76 static volatile int aio_fd_cnt;
77 volatile int __aio_fut;
79 static size_t io_thread_stack_size;
81 #define MAX(a,b) ((a)>(b) ? (a) : (b))
83 static struct aio_queue *__aio_get_queue(int fd, int need)
90 unsigned char b=fd>>16, c=fd>>8, d=fd;
91 struct aio_queue *q = 0;
92 pthread_rwlock_rdlock(&maplock);
93 if ((!map || !map[a] || !map[a][b] || !map[a][b][c] || !(q=map[a][b][c][d])) && need) {
94 pthread_rwlock_unlock(&maplock);
95 if (fcntl(fd, F_GETFD) < 0) return 0;
96 pthread_rwlock_wrlock(&maplock);
97 if (!io_thread_stack_size) {
98 unsigned long val = __getauxval(AT_MINSIGSTKSZ);
99 io_thread_stack_size = MAX(MINSIGSTKSZ+2048, val+512);
101 if (!map) map = calloc(sizeof *map, (-1U/2+1)>>24);
103 if (!map[a]) map[a] = calloc(sizeof **map, 256);
104 if (!map[a]) goto out;
105 if (!map[a][b]) map[a][b] = calloc(sizeof ***map, 256);
106 if (!map[a][b]) goto out;
107 if (!map[a][b][c]) map[a][b][c] = calloc(sizeof ****map, 256);
108 if (!map[a][b][c]) goto out;
109 if (!(q = map[a][b][c][d])) {
110 map[a][b][c][d] = q = calloc(sizeof *****map, 1);
113 pthread_mutex_init(&q->lock, 0);
114 pthread_cond_init(&q->cond, 0);
119 if (q) pthread_mutex_lock(&q->lock);
121 pthread_rwlock_unlock(&maplock);
125 static void __aio_unref_queue(struct aio_queue *q)
129 pthread_mutex_unlock(&q->lock);
133 /* This is potentially the last reference, but a new reference
134 * may arrive since we cannot free the queue object without first
135 * taking the maplock, which requires releasing the queue lock. */
136 pthread_mutex_unlock(&q->lock);
137 pthread_rwlock_wrlock(&maplock);
138 pthread_mutex_lock(&q->lock);
142 unsigned char b=fd>>16, c=fd>>8, d=fd;
145 pthread_rwlock_unlock(&maplock);
146 pthread_mutex_unlock(&q->lock);
150 pthread_rwlock_unlock(&maplock);
151 pthread_mutex_unlock(&q->lock);
155 static void cleanup(void *ctx)
157 struct aio_thread *at = ctx;
158 struct aio_queue *q = at->q;
159 struct aiocb *cb = at->cb;
160 struct sigevent sev = cb->aio_sigevent;
162 /* There are four potential types of waiters we could need to wake:
163 * 1. Callers of aio_cancel/close.
164 * 2. Callers of aio_suspend with a single aiocb.
165 * 3. Callers of aio_suspend with a list.
166 * 4. AIO worker threads waiting for sequenced operations.
167 * Types 1-3 are notified via atomics/futexes, mainly for AS-safety
168 * considerations. Type 4 is notified later via a cond var. */
171 if (a_swap(&at->running, 0) < 0)
172 __wake(&at->running, -1, 1);
173 if (a_swap(&cb->__err, at->err) != EINPROGRESS)
174 __wake(&cb->__err, -1, 1);
175 if (a_swap(&__aio_fut, 0))
176 __wake(&__aio_fut, -1, 1);
178 pthread_mutex_lock(&q->lock);
180 if (at->next) at->next->prev = at->prev;
181 if (at->prev) at->prev->next = at->next;
182 else q->head = at->next;
184 /* Signal aio worker threads waiting for sequenced operations. */
185 pthread_cond_broadcast(&q->cond);
187 __aio_unref_queue(q);
189 if (sev.sigev_notify == SIGEV_SIGNAL) {
191 .si_signo = sev.sigev_signo,
192 .si_value = sev.sigev_value,
193 .si_code = SI_ASYNCIO,
197 __syscall(SYS_rt_sigqueueinfo, si.si_pid, si.si_signo, &si);
199 if (sev.sigev_notify == SIGEV_THREAD) {
200 a_store(&__pthread_self()->cancel, 0);
201 sev.sigev_notify_function(sev.sigev_value);
205 static void *io_thread_func(void *ctx)
207 struct aio_thread at, *p;
209 struct aio_args *args = ctx;
210 struct aiocb *cb = args->cb;
211 int fd = cb->aio_fildes;
213 void *buf = (void *)cb->aio_buf;
214 size_t len = cb->aio_nbytes;
215 off_t off = cb->aio_offset;
217 struct aio_queue *q = args->q;
220 pthread_mutex_lock(&q->lock);
221 sem_post(&args->sem);
228 at.td = __pthread_self();
231 if ((at.next = q->head)) at.next->prev = &at;
235 int seekable = lseek(fd, 0, SEEK_CUR) >= 0;
236 q->seekable = seekable;
237 q->append = !seekable || (fcntl(fd, F_GETFL) & O_APPEND);
241 pthread_cleanup_push(cleanup, &at);
243 /* Wait for sequenced operations. */
244 if (op!=LIO_READ && (op!=LIO_WRITE || q->append)) {
246 for (p=at.next; p && p->op!=LIO_WRITE; p=p->next);
248 pthread_cond_wait(&q->cond, &q->lock);
252 pthread_mutex_unlock(&q->lock);
256 ret = q->append ? write(fd, buf, len) : pwrite(fd, buf, len, off);
259 ret = !q->seekable ? read(fd, buf, len) : pread(fd, buf, len, off);
269 at.err = ret<0 ? errno : 0;
271 pthread_cleanup_pop(1);
276 static int submit(struct aiocb *cb, int op)
280 sigset_t allmask, origmask;
282 struct aio_queue *q = __aio_get_queue(cb->aio_fildes, 1);
283 struct aio_args args = { .cb = cb, .op = op, .q = q };
284 sem_init(&args.sem, 0, 0);
287 if (errno != EBADF) errno = EAGAIN;
293 pthread_mutex_unlock(&q->lock);
295 if (cb->aio_sigevent.sigev_notify == SIGEV_THREAD) {
296 if (cb->aio_sigevent.sigev_notify_attributes)
297 a = *cb->aio_sigevent.sigev_notify_attributes;
299 pthread_attr_init(&a);
301 pthread_attr_init(&a);
302 pthread_attr_setstacksize(&a, io_thread_stack_size);
303 pthread_attr_setguardsize(&a, 0);
305 pthread_attr_setdetachstate(&a, PTHREAD_CREATE_DETACHED);
306 sigfillset(&allmask);
307 pthread_sigmask(SIG_BLOCK, &allmask, &origmask);
308 cb->__err = EINPROGRESS;
309 if (pthread_create(&td, &a, io_thread_func, &args)) {
310 pthread_mutex_lock(&q->lock);
311 __aio_unref_queue(q);
312 cb->__err = errno = EAGAIN;
313 cb->__ret = ret = -1;
315 pthread_sigmask(SIG_SETMASK, &origmask, 0);
318 while (sem_wait(&args.sem));
324 int aio_read(struct aiocb *cb)
326 return submit(cb, LIO_READ);
329 int aio_write(struct aiocb *cb)
331 return submit(cb, LIO_WRITE);
334 int aio_fsync(int op, struct aiocb *cb)
336 if (op != O_SYNC && op != O_DSYNC) {
340 return submit(cb, op);
343 ssize_t aio_return(struct aiocb *cb)
348 int aio_error(const struct aiocb *cb)
351 return cb->__err & 0x7fffffff;
354 int aio_cancel(int fd, struct aiocb *cb)
356 sigset_t allmask, origmask;
357 int ret = AIO_ALLDONE;
358 struct aio_thread *p;
361 /* Unspecified behavior case. Report an error. */
362 if (cb && fd != cb->aio_fildes) {
367 sigfillset(&allmask);
368 pthread_sigmask(SIG_BLOCK, &allmask, &origmask);
371 if (!(q = __aio_get_queue(fd, 0))) {
372 if (errno == EBADF) ret = -1;
376 for (p = q->head; p; p = p->next) {
377 if (cb && cb != p->cb) continue;
378 /* Transition target from running to running-with-waiters */
379 if (a_cas(&p->running, 1, -1)) {
380 pthread_cancel(p->td);
381 __wait(&p->running, 0, -1, 1);
382 if (p->err == ECANCELED) ret = AIO_CANCELED;
386 pthread_mutex_unlock(&q->lock);
388 pthread_sigmask(SIG_SETMASK, &origmask, 0);
392 int __aio_close(int fd)
395 if (aio_fd_cnt) aio_cancel(fd, 0);
399 void __aio_atfork(int who)
402 pthread_rwlock_rdlock(&maplock);
405 if (who>0 && map) for (int a=0; a<(-1U/2+1)>>24; a++)
406 if (map[a]) for (int b=0; b<256; b++)
407 if (map[a][b]) for (int c=0; c<256; c++)
408 if (map[a][b][c]) for (int d=0; d<256; d++)
410 pthread_rwlock_unlock(&maplock);
413 weak_alias(aio_cancel, aio_cancel64);
414 weak_alias(aio_error, aio_error64);
415 weak_alias(aio_fsync, aio_fsync64);
416 weak_alias(aio_read, aio_read64);
417 weak_alias(aio_write, aio_write64);
418 weak_alias(aio_return, aio_return64);